Showing papers by "Eric R. Heller published in 2015"
TL;DR: Finite element simulation results of the transient stress response of an AlGaN/GaN high electron mobility transistor (HEMT) suggest transient failure mechanisms may differ from those previously studied under DC operation due to large amount of cyclic loading of a device around the gate structure.
34 citations
TL;DR: In this article, a very high mobility Debye tail was observed in Ga-doped ZnO films of thicknesses d = 5, 25, 50, and 300 nm.
Abstract: Highly Ga-doped ZnO (GZO) films of thicknesses d = 5, 25, 50, and 300 nm, grown on 160-nm ZnO buffer layers by molecular beam epitaxy, had 294-K Hall-effect mobilities μH of 64.1, 43.4, 37.0, and 34.2 cm2/V-s, respectively. This extremely unusual ordering of μH vs d is explained by the existence of a very high-mobility Debye tail in the ZnO, arising from the large Fermi-level mismatch between the GZO and the ZnO. Scattering theory in conjunction with Poisson analysis predicts a Debye-tail mobility of 206 cm2/V-s at the interface (z = d), falling to 58 cm2/V-s at z = d + 2 nm. Excellent fits to μH vs d and sheet concentration ns vs d are obtained with no adjustable parameters.
12 citations
TL;DR: In this article, the surface potential of the operating device was mapped in cross-section by KPFM, effectively mapping the device in two-dimensional cross-sectional, including under metallization layers (i.e., gate, field plates and ohmic contacts).
Abstract: AlGaN/GaN high-electron-mobility transistors (HEMTs) were characterized in cross-section by Kelvin probe force microscopy (KPFM) during in situ biasing. The HEMTs used in this study were specially designed to maintain full and representative transistor functionality after cross-sectioning perpendicular to the gate width dimension to expose the active channel from source to drain. A cross-sectioning procedure was established that produces samples with high-quality surfaces and minimal degradation in initial transistor performance. A detailed description of the cross-sectioning procedure is provided. Samples were characterized by KPFM, effectively mapping the surface potential of the device in two-dimensional cross-section, including under metallization layers (i.e., gate, field plates, and ohmic contacts). Under the gate and field plate layers are where electric field, temperature, and temperature gradients are all most commonly predicted to have peak values, and where degradation and failure are most likely, and so this is where direct measurements are most critical. In this work, the surface potential of the operating device was mapped in cross-section by KPFM. Charge redistribution was observed during and after biasing, and the surface potential was seen to decay with time back to the prebias condition. This work is a first step toward directly mapping and localizing the steady-state and transient charge distribution due to point defects (traps) before, during, and after device operation, including normally inaccessible regions such as under metallization layers. Such measurements have not previously been demonstrated for GaN HEMT technology.
10 citations
TL;DR: The results demonstrate the importance of discriminating between traps throughout the entire bandgap with regard to the relative roles of individual traps on degradation of GaN HEMTs after ALT.
4 citations
TL;DR: Little permanent degradation is observed due to hot electron effects in GaN HEMTs at these extreme operating conditions and it is inferred that other considerations, such as key dimensions in channel or peak electric field, are more relevant to physics of failure than drain bias alone.
3 citations